There’s some movement in the Oracle-Google lawsuit today, but it’s rather difficult to determine just what kind of movement. The jury was told by the judge Alsup to assume APIs are copyrightable – something Alsup still has to determine later during trial – and with that in mind, the judge ruled Google violated Oracle’s copyright on Java. However, the jury did not come to an agreement on a rather crucial question: whether or not it was fair use. All in all, a rather meaningless verdict at this point, since it’s incomplete. Also, what kind of nonsense is it for a judge to tell a jury to assume something is illegal? Am I the only one who thinks that’s just complete insanity?
No, you’re not the only one who thinks that’s just complete insanity
“Also, what kind of nonsense is it for a judge to tell a jury to assume something is illegal? Am I the only one who thinks that’s just complete insanity?”
Thom…you really should try and get some vague kind of information before mouthing off on your blog, I’m sorry, serious news site.
It’s not nonsense, it’s how many legal systems work.
‘Whether the APIs are copyrightable’ is a question of law. Questions of law are for the judge, not the jury, to decide. Therefore the judge didn’t ask the jury that; this is quite correct. ‘Whether, if the APIs are copyrightable, Google infringed that copyright’ is a question of fact. Questions of fact, not questions of law, are what juries decide, in the American system (and most jury systems).
So the judge asked the jury to answer the question of fact, and will himself answer the question of law. That’s exactly how the system is supposed to work. It’s not insane at all.
I find a jury system insane enough even without nonsense like this, but alas. I’m glad we don’t have this medieval nonsense in this country.
But that doesn’t negate the fact that the judge asked the jury to render a verdict on something we don’t even know is illegal in the first place. For someone from a country without jury trial and based entirely on statutory law, this IS pure insanity.
Look, don’t take it out on me we don’t report on every Fedora fart anymore (I don’t know what else could have crawled up your butt recently).
Edited 2012-05-07 20:33 UTC
Yes your system functions way better:
http://www.dailymail.co.uk/travel/article-2061730/Dutch-cannabis-co…
Thinking about this… I know exactly why this is so idiotic.
If not even a judge knows if APIs are copyrightable or not, how should Google have known?
A rather key element of a proper judicial system is that you cannot be found guilty of something that was not illegal at the time you did that something. E.g., if the legal limit is 80kph and I drive 80kph on Monday, and the limit is lowered on Tuesday to 70kph, I cannot be fined for my “speeding” the day before.
This is the same. Apparently, nobody knows if APIs can be copyrightable. So, how could Google have known?
To me, from the perspective of my legal system, this is pure insanity, because it creates a hell of a lot of uncertainty for civilians and companies alike.
Edited 2012-05-07 21:12 UTC
Very insightful point!
Not that I disagree with your point on an intellectual level, but this is a civil law suit. Things being “illegal” has little to do with it – the purpose of civil law is to right wrongs and settle disputes. You do not have to actually break any laws to wrong someone.
It would be possible, for example, for the court to determine that APIs are in fact not copyrightable and still award damages to Oracle if they feel that Google acted in bad faith or otherwise unfairly took advantage of Sun/Oracle.
I’m not saying I think Google did anything wrong – I actually don’t. I’m just saying the question of whether or not APIs are copyrightable, while being very significant to most of us, is not necessarily the only issue nor should it be.
The judge instructing the jury as was done is quite common. It allows the system to short circuit (like an if statement in programming). If the jury finds that Google did not violate the Oracles “assumed” API copyright, well then the matter of whether or not APIs are copyrightable is no longer relevant and everyone can go home…
Yes, but even in civil cases in The Netherlands, you have to have actually done something wrong. You have to have broken the our civil code (not sure how to translate it). It’s distinctly different from the way common law works.
So, just to reiterate: for someone from The Netherlands without common law this just seems incredibly unfair. I just don’t understand how you can be found guilty of something of which it has not even been established if it is illegal at all!
Well, I’m not a lawyer but I’d assume it’s a way of multitasking.
To have an actual violation, we need to determine:
(1) is API copyrightable?
(2) if API is copyrightable, does Google violate it?
Or put in logic:
(1) A is true?
(2) (A => B) is true?
The jury concluded that implication (2) is valid, but that only means something when A is true. So once it’s ruled that API’s are in fact copyrightable then Google violated it and a new jury decision will not be needed.
Or it may be some legal trick I’m not aware of, it just seemed the most logical explanation to me.
It would be an interesting scenario where software is not a derived work in one continent but is considered as one in another. “This software may not be exported to Cuba, Iran, Libya, North Korea, Sudan, Syria and the United States of America.”
Why go over (2) if A is not valid,
That’s like first executing the body and then looking at the if-statement to dismiss the work you put into the body if it’s a 0. Quite ineffective if you ask me. Never got the whole common law thing anyway, I know it allows the system to be more nimble, and to adapt more easily to knew situations. It is also more error prone, and it puts power into peoples’ hands that should not have them. A judge and jury are no politicians who should be making the law. The legal system provides for people that establish whether or not a law has been broken. Separation of power is what it’s called.
Because the judge was hoping (2) would be proven false. Then he would not need to decide on (1) which is a much harder thing.
As another poster mentioned, this happens all the time. Most processors with large pipelines try to predict branches and execute the next instructions in one branch before the if-check is even finished. If it was the right branch, we just saved some time. If it was the wrong one, well just throw those temporary results away and start on the other branch. It beats doing nothing at all. This make sense because some branches are much more likely to occur than others (think while-loops).
Thus it’s better to do something and hope it might be useful than doing nothing while waiting for some results. The jury had to make a few decisions anyway, might as well give them the extra question if that saves us some time later.
Edited 2012-05-09 02:15 UTC
It does seem, on the face of it, a waste of time and perverse for a jury to adjudicate on a matter of fact on the assumption it’s illegal when the judge has yet to give a judgement on the law.
If they’d found against would that mean the judge wouldn’t have bothered with the judgement on law? Presumably, more important for everyone else if the judgement did or didn’t open yet another Pandora’s box of future litigation.
If they’d found against would that mean the judge wouldn’t have bothered with the judgement on law?
Bingo! The judge doesn’t want to rule and potentially set any precedent of law unless it’s required.
edit: My guess is the judge was thinking (hoping) the jury would find there was no copyright infringement or it was “fair use” or “de minimis”.
Edited 2012-05-07 23:06 UTC
Don’t sweat it Thom. The Jury has only decided that IF the API’s are subject to copyright, then Google copied them. The Jury does bot get to say if the APIs are, or are not, subject to copyright.
Here is a source indicating Oracle’s and Google’s reactions to the findings:
http://www.groklaw.net/article.php?story=20120507122749740
Partial Verdict; Oracle Wins Nothing That Matters
Recap of the day: Google won everything but the one issue that the judge has to decide anyway, the API SSO issue. The jury found, as they had been instructed to assume for the purposes of deliberation, that APIs can be copyrighted, the structure, sequence and arrangement of APIs, but that is by no means established. The same question, in a b) section, asked if fair use excused any infringement if found, and the jury couldn’t resolve that issue. But the judge has to decide whether or not that is true, that APIs can be protected by copyright. That comes later this month. Meanwhile, Oracle prevailed only on 9 lines of code that Google admitted prior to trial to have included by mistake and then removed from current Android. Oracle’s own expert, the judge pointed out in court, valued those 9 lines of code at zero. This is 9 lines out of millions. So that means, if we are looking at damages, that so far Oracle has won nothing. There is no liability. You can’t have infringement without considering fair use, Google asserts, and there will be briefing on that. Somebody has to decide that fair use issue. And then the judge has to decide about the API copyrightability issue. If he rules that APIs can’t be copyrighted, as the EU Court of Justice just ruled, then fair use is moot. And Oracle takes nothing at all from the copyright phase of this litigation, and this was heralded far and wide by Oracle people as the big ticket item, if you recall.
Don’t let anyone fool you. Today was a major victory for Google. That’s why after the jury left, our reporter says that Google’s table was laughing, and Oracle’s mighty glum.
People seem constantly focused on the wrong problem. The question is: can an API contain elements which are uniquely creative and expressive? If so, those expressive elements can be copyrighted.
The judge doesn’t have a problem knowing whether or not APIs can be covered by copyrights; he has difficulty disassembling what is merely functional and what is expressive.
While not strictly analogous, similar complexities can be seen when a movie “infringes” a novel by changing character names and a few plot points, changing all of the dialogue but containing a majority of the conversations expressing largely the same point, purpose, values, and content. The judge doesn’t have any difficulty knowing that a novel can be copyrighted, but determining what is the uniquely expressive qualities of the source material and if the derivative relies on the source for an unfair balance of its own expressive qualities can be a substantially complex difficulty.
Edited 2012-05-08 00:46 UTC
The world changes every day creating exciting new ways to bring harm to others. If everything had to be explicitly pre-codified, there would be a lot of unaddressed injustice in the world.
Most legal systems were designed to evolve. If somebody acts in an “unreasonable” way or performs an action that signicantly works against the common good then they can be judged to have acted illegally.
The US system strives to avoid creating new laws unless they are both a natural extension to precedent and necessary to support a verdict of infringement by a jury.
It actually does make sense. The question of whether or not existing copyright law applies to APIs is a matter of interpreting the law. This is a task for a judge, and not a jury, so jury was instructed to assume that copying APIs is a copyright violation, so that way their only question was whether or not Google copied the APIs.
Now that the jury decided that Google did copy the APIs, it is up to the judge to decide whether or not APIs are actually covered by copyright. It is possible for the judge to rule that APIs are not eligible for copyright, in which case the jury’s decision is irrelevant.
Or, perhaps there is existing precedent somewhere that indicates that APIs are copyrightable. While that seems unlikely at this point, if this were the case, should Google still be not liable because “they didn’t know?” In this situation, I would argue no.
Finally, if there is no precedent, and the judge decides that APIs are copyrightable, should Google still be liable, even though the law is unclear? I would argue that, yes, they are. To do otherwise would deny Oracle protection under the law. While Google could argue that the law was unclear and they shouldn’t be made to pay, Oracle can just as easily make the argument that the law was clear enough for them. You can’t fault Oracle for taking the position if the judge rules in a way favorable to them. In this case, it is the fault of the legislators for failing to write clear laws or to adapt.
Again, while it seems unfair that Google should be made to pay for unclear legislation, it is arguably more unfair that Oracle shouldn’t be protected.
Either way, Judge Alsup indicated that he may rule in Google’s favor on the issue of whether or not APIs are copyrightable. From Ars Technica’s writeup on the verdict: “But at the end of the trial, Alsup gave the lawyers a short talk indicating that he may see the API issue Google’s way, comparing APIs to an idea for writing a guide book from San Francisco to Monterey.” ( http://arstechnica.com/tech-policy/news/2012/05/jury-rules-google-v… )
I do not live in the US but Thom’s reaction is mostly one of ignorance here.
1) The judge will avoid creating new law unless he has to
2) The judge will only decide if APIs can be copyrighted if Google is found to have done that
3) The jury is assumed to be impartial, unbiased, and to be a source of common sense
4) The jury will decide if Google has wrongfully copied the API
5) In order to get a ruling from the jury, the judge has to tell them to assume it is illegal
6) The judge is assumed to be capable of forming an informed legal opinion on whether or not an activity is illegal
7) If Google is found to have infringed, the judge will decide if what they did was illegal
8) Instead of relying on the (possibly biased) decision of one man (the judge) when assessing guilt the US relies on the consensus of twelve people (selected to be unbiased)
9) When deciding complex legal technicalities, the US relies on the judgement of one highly educated and experienced person (the judge) instead of the purposely uneducated opinions of a group (the jury)
10) The whole thing is subject to appeal
It is cumbersome, but it actually makes a fair bit of sense to me.
Juries decide “did” or “did not”.
Judges decide “legal” or “illegal”
Yeah, I get what you are saying. Its often that cases that do adress unclear areas of the law are thrown out on an unrelated matter. Judges won’t often make a ruling on an unclear part of the law unless they have to. It prevents arguing of hypothetical situations that will never happen in real life.
Why would the judge have to make a new law, if it’s not illegal, it’s legal. If there needs to be a new law, it’s up to politicians to create them. These new laws are triggered by unfair results in trials, granted, you loose some effectivity there, but it avoids the result of the fairness of a trial to depend on one (or a few) mans'(mens’) vision(s). A judge only has to decide whether or not it is illegal what has been done. That’s where the common law, to me, (and to thom) seems so flawed. The seggregation of power is really not respected in common law and has very nasty unwanted sideeffects that are not negated by the advantage of a nimble judicial system.
Heh. Try to make out I have some kind of agenda. Cute, but not going to wash. Sorry, try a different number.
It’s a simple matter of efficiency. The jury has to decide one thing. The judge has to decide another thing. Should they a) try to get both things decided as soon as possible or b) do one and then the other, even if it’s perfectly possible to do both at once? I’m pretty sure you wouldn’t _really_ prefer b). Trials take long enough already. Hell, it’s just parallellization. (edit: the point that the judge didn’t want to create ‘new law’ by ruling on whether APIs are copyrightable unless he actually _had_ to is also an important one – probably more important).
Trials are messy affairs where things like this happen all the time. If the media didn’t jump on intermediate hearings and determinations like they were the Word of God, less of everyone’s time would be wasted.
edit: as far as ‘no-one’s sure if it’s illegal’ goes, well, that happens under just about any code, really. It’s incredibly difficult to write a law that requires absolutely no kind of interpretation. “Thou shalt not kill” – fine – but what about self-defence? What if you’re insane? What if you didn’t really mean to? Some kinds of interpretation are done by having the legislators traipse back in and refine the law, sure, but some are done by judges. In this case, the question of whether APIs constitute copyrightable material as defined in the U.S. copyright law is a question of interpretation of the relevant law, and it’s a judge’s job to perform that interpretation.
Edited 2012-05-08 07:18 UTC
Are you glad you have the inherently flawed Dutch system? Just wondering.
I was involved in 3 court cases, one included a charge of hacking, and 3 times I came out a winner so this Dutch person is quite happy with the system!
Put separately – what’s crawled up my butt lately is simply the declining quality of the site. Sorry, but that’s the truth. I guess instead of snarking I should just take it out of my RSS feed, really. I don’t recall the last time I found out something interesting from OS News.
One could blame Thom for posting news (and adding a layer of bias) that can be found on x other sites, but on the other hand these articles seem to get a lot more attention from the readers than the non-mainstream ones.
Thus it seems Thom is providing what the majority of readers want, so in that regard he and his/this site are doing fine. For a job he doesn’t get paid for he puts in a lot of time and effort.
It’s not what you (and I) want to read and you (and I) long for the way this site was in the past, always hoping one day it will return to that state of joyfulness, but I my doubt is ever increasing.
Funny enough, you never complain when the bias is in Apple’s favour. Funny how that works, eh
.
I pick my moments to turn a blind eye with laser cunning precision and unparalleled Vulcan logic.
I think there’s a kind of ‘lag’ effect. I’m more likely to comment on big controversial stories, but at the same time, I’m thinking ‘man, I’m wasting time and effort on this’. And eventually I’ll just cancel my subscription. Every so often I go through my feeds and do that. So the effect is kinda delayed, which makes it less obvious.
I’m less likely, actually, to comment an interesting story that teaches me something new than I am on an inaccurate story about something I already understand. After all, I have nothing useful to contribute on the first; but I do on the second. But I’m more likely to keep reading a site that posts the first than I am a site that posts the second. So I don’t think ‘number of comments’ is a reliable proxy metric for ‘value of site to readers’, necessarily.
You are right, but I also think that the comments section adds value to this site. Most sites I come across have crappy comments sections.
I probaly spend more time reading comments than I do reading articles. Often I just read the headline/intro and skip to the comments.
…which happens to be the reason I always *try* to get the main story link in the teaser. It doesn’t always work, but it does, most of the time. I do this exactly because nobody ought to be forced to read my crap.
You’re clever, but not clever enough to spot me drive by Alkmaar on my way to Bergen.
I don’t think you’re correct with this conclusion. Every programmer knows “if – then – else” statements, conditionals. They consist of a condition and a conclusion. In this particular case (if I understood it correctly!), the condition is an assumption, a possibility, a definition, or a theory. There is no evidence (yet) that this condition is true.
“If the moon consists of green cheese, would Google have infringed that copyright?”
The conclusion (as we don’t know if the condition is true or not) may be even useless. See the rules for implications, which are the logical equivalent of a conditional for “condition x conclusion = truth of statement”:
true x true = true
true x false = false
false x true = true
false x false = true
How does logic match in the field of law? Can argumenting on a “what if” basis be applied here? How does discussing possibilities (and evaluating them by the jury, with the judge setting up the preconditions) affect the legal process?
Not familiar with out-of-order execution and branch prediction are we?
It works something like this:
The code:
law = Law(API | COPYRIGHTABLE);
if (law.IsIllegal(google->DoocumentedUseCase()))
jury->RenderVerdict();
We know that the if statement can only have two possible outcomes: true or false. Further, we know that if the answer if false, the jury->RenderVerdict() is unneeded.
We find no basis in which we should assume false, so we assume true until we can calculate law.IsIllegal()’s return value and so we throw jury->RenderVerdict() into the pipeline. Now, should we get around to a false result from law.IsIllegal() we can cancel the execution of jury->RenderVerdict() and also completely free all related branches and predictions. Now, if jury->RenderVerdict() returns before we have completed law.IsIllegal(), we need only to cache the result and further execution in regards to jury can occur.
This makes things much faster.
Hope that makes sense…
–The loon
LMAO at your use of programming logic to explain the situation. I would love to see how a judge (or anyone familiar with the justice system) would react on seeing that!
Oh, some legal systems really are out of order, sometimes there even should be an execution, and predicting how branches would behave means entering the magical realm of dreaming fairies and wonder bunnies. This means: You cannot beat law with logic, because it’s often free of any common sense, just like my comment. 🙂
I hope you’ve got a license to use predicate logic from Mr Boole – it looks like you might be infringing to me. Can’t you use something more original the ‘true’ and ‘false’?
shouldn’t it be
true x true = true
true x false = false
false x true = false
false x false = true
?
No, for the implication, true x false = false is correct. Read: “from something true, a false statement cannot emerge”; and false x true = true is also correct, because “from something false, a true statement can emerge”. That’s the implication, it’s not the “logical and” or “logical xor” (in which case your commented statement would be correct). Implication means “if A then B” (formal: “A implies B”), and the whole construct (not only A and B) can have a truth value. More formal: A -> B <=> -A v B (read “A implies B when not A or B”), and you can put in true and false for A and B and check for all 4 cases.
However, law isn’t logic. From something plain stupid, a ridiculous ruling can always emerge. 🙂
http://www.groklaw.net/article.php?story=20120507122749740
tl;dr.wec: Good for Google
Hi,
For something like “if( (condition1) && (condition2) ) { return GUILTY; } else { return INNOCENT; }” where “condition1” is hard to figure out and “condition2” is easy to figure out; it is reasonable to determine if “condition2” is false first. If you’re lucky and you may be able to return “INNOCENT” without ever evaluating “condition1” at all (and without setting a precedent when you’d rather avoid doing so).
– Brendan
the difference is that the easy part that could have been avoided is a repetitive task. The difficult part has to be done anyway. If not it will have to be answered for the next case that comes along, unless of course the same logic is followed. Following your logic it could result in a lot of lawsuits where the jury first finds that the parties did not infringe on the copyright of someone elses API. And that it is only after the hundred and first lawsuit it will get an answer on whether or not it is illegal in the first place. Answering the hard question has to be done only once, and it would be good to actually know the answer.
thats interesting
i thought interoperability-related stuff (and imo APIs are exactly that) is one of the special-cases where even the dmca doesn’t apply
another think that i just remembered:
3dfx sued nvidia for reimplementing glide ond their cards (must have been tnt2)
and if i remember correct nvidia wasn’t found guilty by the judge
Although I have little regard for the American legal system, I do not see the whole APIs being copyrightable thing getting through this trial, let alone all the way through to the supreme court.
However, if I am wrong that even if this turns out to be the case, it is going to come back to bite Oracle hard in the ass.
How many 3rd party API do they use in their software? How many of the APIs in Solaris or their DBs were created by or properly owed/licensed by them?
Much like oracles of old, this company seems to be full of fume huffing fools who have no particular insight into the future.
IANAL, but I think this is what is going on:
In the American legal system, the jury is responsible for deciding factual matters while the judges interpret the law. However, judges are only supposed to rule on cases brought before them.
So the jury was basically asked “assuming APIs are copyrightable, was there copying”? They said yes. They were then told the applicable rules for “fair use” and asked if those applied. They said “we can’t decide”.
To convict Google, Oracle would need both the jury and the judges to fill in their parts.
I caught two brief CNBC reports on this. Both times I just knew they wouldn’t mention the fact that the judge would be deciding whether or not APIs are copyrightable. And they didn’t disappoint me.
Anyway, buried in one of the threads at Groklaw someone mentions, along with the main point that others have made here concerning matters of law and matters of fact, that the Judge is going this route in anticipation of appeals.
If it turns out that APIs are copyrightable, wouldn’t this by the precedent also make e.g. Wine and ReactOS breaking Microsoft’s copyrights?
Of course, a judge cannot order a jury to deliver any verdict, if they cannot figure it out, then its a mistrial and we start again. Perhaps the judge wants this to go to appeal where it will certainly be overturned.
“Of course, a judge cannot order a jury to deliver any verdict”
Actually, judges can direct juries to acquit, in certain circumstances.